1 Introduction

Prairie dogs, a species of ground squirrel who live in tunnels under the ground, have developed a complex communication system. When an intruder enters their territory, they do not only tell each other whether it’s a human, a dog or someone from another species, they also describe this intruder in detail. In the case of a human, they for example mention their height, the colour of their hair and T-shirt, and whether or not they carry an object, such as for example an umbrella or a gun (Slobodchikoff et al. 2009). Prairie dogs are not the only nonhuman animals who have more elaborate systems of communication than humans have always thought. What is exceptional is that their language has been studied in this much detail.

For a long time, nonhuman animal capacities were mostly studied to better understand humans (Meijer 2019). This is changing. Recent years have seen a turn towards studying nonhuman animals’ languages, cultures, emotional lives, cognitive capacities and even politics (Meijer 2019; Bekoff and Pierce 2009; Brosnan and De Waal 2012; Leavens 2007; Wasserman and Zentall 2012). These studies show their inner lives are more complex than previously assumed, and that interspecies communication about many issues is possible, raising questions about the relationships between humans and other animals. Humans can no longer assume they know best—other animals have perspectives on their lives, and on their relations with humans, too—or treat nonhuman animals only as objects of study. In order to formulate what is just, ethically or politically, we therefore need to engage differently with them. The first step in this process is to recognize that their agency matters. Similar to humans, other animals have an interest in shaping their own lives. The second step is to get a better understanding of how different animals express themselves, their desires and views, in order to be able to build new, and better, relations with them.

Taking into account animal agency in ethics and politics is however not easy. Current views about nonhuman animals and their agency are shaped by stereotypical or anthropocentric ideas about their capacities, that were formed within a behaviouristic framework. If we believe that we cannot get to know anything about what goes on in nonhuman animals’ minds from watching their behaviour or engaging with them, it will be difficult to recognize animal agency in the first place. Furthermore, human treatment of other animals is often focused on curtailing their agency. In farming practices this not only means using fences and other material devices to limit their freedom of movement, but also modifications of their bodies (such as the cutting of beaks and tails) and even genetic interventions. Our stereotypical views on animals and the curtailing of their agency limit our understanding of animals’ capacity for agency. Both these aspects imply that humans currently should practice epistemic humility—there is much we do not know about other animals, and much of what we think we do know is formed by stereotypical views. Much scientific knowledge about nonhuman animal capacities reflects this. Our questions determine the answers other animals can give, and we long asked the wrong questions (Despret 2016). For example, if we want to know whether animals can make conscious choices, this will be difficult to find out when the animals are kept in a setting that is devoid of possibilities for making choices. Getting to know more about animals thus requires a critical investigation of existing concepts and knowledge, aimed at decentring the human, and new forms of animal research.

However, we also cannot refrain from considering the question of animal agency when thinking about ethics. We currently live in an epoch in which human actions curtail the agency of animals, and the ways they can shape their own lives, more than ever before. We cannot simply let them be, human and animal lives are thoroughly intertwined. Due to human impact on ecosystems, nonhuman animals and the planet, our current age has been named the Anthropocene: the age of the human. The Anthropocene has shaped the knowledge and technology for humans to realize that animals have more agency than has been assumed, but ironically it is also an epoch where animal agency is increasingly curtailed. Captive animals, for example, are curtailed in their agency when their enclosure is too small or they cannot go out to look for new experiences. This means they cannot express their ‘inquisitive exploration’ and when they are solitarily housed they cannot ‘engage in social play’ (Špinka and Wemelsfelder 2011, 36). They will not acquire certain competences and their level of interaction and self-expression will be limited. In contrast, in the wild they are confronted more often with novel challenges for which they have to seek a solution and this stimulates their agency, as well as opportunities for building social relations and moving around. Animals in the wild, however, are increasingly curtailed in their agency as well, due to habitat loss and fragmentation and changing climatic conditions. These conditions lead to new conflicts between humans and other animals, and problems of justice.

In this chapter we explore the questions of animal ethics and animal agency in tandem, because they are interconnected: taking seriously animal agency has ethical consequences and animal ethics should take into account animal agency. Furthermore, as the concept agency currently is used in many different ways in different fields, we also aim to shed light on the concept itself, focusing on its meaning in philosophy. Our focus of inquiry is not animals’ moral agency, which we see as part of the larger spectrum of agency; some animals may well be moral agents and others perhaps not. We instead focus on the ethical consequences of the fact that other animals are actors with their own perspective on life, and on relations. Moreover, we do not take animal agency as an ‘entry ticket’ for the moral community, or as synonymous with moral status. We assume that all animals that have subjective experiences should be attributed moral status and our point is that it is in the interest of all beings with moral status to have their agency taken seriously, to the extent that they have agency.

We begin by reviewing existing approaches to animal agency, and discuss their shortcomings and strengths, formulating a working definition of agency. We then argue for a relational model of ethics that takes animal agency seriously at the micro- and macro level. We end by discussing initial steps towards formulating new relations with other animals. As we do not know the precise scope of animal agency yet, given that humans have so long ignored it, and even oppressed it in so many ways, we do not aim to write a final statement, but we rather see this as an investigation into developing a new, relational, animal ethics for the Anthropocene.

2 Conceptualizing Animal Agency: Two Models

Animal agency is currently used as an umbrella term for ways in which animals act and influence the world around them in various directions of study, and the meaning varies between different fields. In this chapter we focus on its philosophical meaning. More specifically, in this section we review two models of conceptualizing agency in relation to animal ethics. The models can be seen as two opposites on a spectrum of approaches. The first we will call propositional agency, following Sebo (2017), in which a specific form of human rationality takes centre stage, leading to anthropocentrism. The second model argues agency is found in everything that has the capacity to move something else, leading to problems for ethical and political theory (Bennett 2010). These two extreme models obviously do not provide a comprehensive overview of all theories on (animal) agency. However, they do bring to light some of the key questions in thinking about nonhuman and human agency, and they show how views of morality have been linked to views about agency in the philosophical tradition.

2.1 Propositional Agency

The concept agency has in philosophy traditionally been reserved for intentional human action and is linked to the capacity for propositional thought (Sebo 2017, see also Bermúdez 2007). In an insightful article about agency and moral status, Sebo (2017) calls this conception of agency ‘propositional agency’. Propositional agency starts from the common sense idea that there is a difference between action and mere behaviour. This difference is in the philosophical tradition often interpreted narrowly: as the difference between intentional action and mere behaviour. Intentional action is specified as acting on ‘judgments about what we have reason to believe, desire, and/or do’ (Sebo 2017, 14). Being capable of intentional action thus formulated presupposes cognitive capacities that other animals were long thought not to possess, such as for example second order thoughts.

Many animal philosophers today challenge this interpretation of agency, together with the underlying view of animal subjectivity (Sebo 2017). They argue that at least some nonhuman animal species possess (some of) these capacities and that differences between humans and other animals in this regard are a matter of degree, and not kind (see for example Gennaro 2009). Furthermore, humans often act habitually (Sebo 2017, Donaldson and Kymlicka 2011) and an image of the human as primarily a rational being relies on an idealized view of humans. Introducing agency as intentional agency in the narrow description above does not do justice to other animals and exaggerates the gap between humans and other animals.

In addition to these problems, propositional agency is also based on specific forms of human reasoning, and values these more than other forms of reasoning. While recent research finds that certain other animals are perhaps capable of these forms of reasoning, this does not do justice to the fact that different species have different forms of agency. Dog agency should be understood as dog agency, not as lesser-than-human agency. Current anthropocentric conceptions of agency can function as a starting point for evaluating others and interpreting relations, but they cannot be an end point. For assessing their epistemic value in multispecies contexts, we need to investigate the power relations that led to current formulations, take into account new empirical research about other animals (including insights from narrative ethology and case studies), and engage with them differently in order to foster their agency instead of constraining it (see Blattner et al. (2020) for a longer discussion, see also Calarco 2018). More generally, new multispecies definitions of concepts—such as agency—should not be based on how much the other animals resemble humans, but include respect for their forms of expression and knowledge formation.

2.2 Materialist Agency

On the complete other side of the spectrum of propositional agency, we find object-oriented theories. One of their proponents, Jane Bennett, formulates a political ecology, in which nonhumans exercise agency on a spectrum with humans. She argues for a ‘vital materialism’ (2010, 23) in which objects possess power, and agency is located in ‘assemblages’: ad hoc groupings of diverse elements that can consist of human and non-human bodies. Bodies are always part of larger networks, which Bennett envisions as webs, or ‘knotted worlds’ of vibrant matter (2010, 13). Objects, or non-humans, the terms are used interchangeably, are interconnected with human bodies, which are themselves made of matter and influenced by pressure from the outside world. To conceptualize the pressure different bodies exercise—or, in other words: their agency—Bennett uses Spinoza’s term ‘conatus’, which means a trending tendency to persist. According to Bennett and Spinoza, non-human bodies share this conative nature with human bodies. Bennett sees these bodies as associative, or even social (2010, 21), in the sense that each body by nature continuously affects and is affected by other bodies.

While Bennett rightly recognizes that agency can be exercised by different beings in different ways, and that specific forms of agency come into being in and through relations, this approach makes normative judgments difficult. Subjective agency dissolves when we are all just bodies moving, and intentions do not seem to matter anymore, which is counterintuitive. Furthermore, the category ‘nonhuman’ runs the risk of reinforcing stereotypical ideas about nonhuman animals, who are grouped with things, and as ever contrasted with human (as in the word ‘nonhuman’). Agency and subjectivity are highly problematic and exclusionary concepts in the philosophical tradition, but reformulating them in this way does not get to the root of the problem, and creates new problems for thinking about nonhuman animal agency. These problems also matter with regard to human responsibility. In an age in which humans dominate the lives of animals of so many other species, taking responsibility is an ethical priority. While Bennett argues that her work is meant to promote this type of responsibility (2010, Introduction), from within the theory it seems hard to realize.

2.3 A Working Definition of Agency

Both the propositional and the materialist approach fail to adequately take into account nonhuman animals’ perspectives. The first because human forms of rationality and agency are taken as the standard, which excludes many other animals (and some humans) beforehand, and the second because it fails to offer a framework in which we can make normative or political judgments about others, and does not explicate new forms of engagement with nonhuman animals, instead grouping them with things. How then, should we understand (animal) agency? We do not want to give a fully developed definition from the outset, as we think an understanding of animal agency should come about in the diverse practices of human-nonhuman animal relationships, and more research needs to be done into such relationships. However, we need a loose working definition of agency in order to give our search some direction.

Irvine (2004) defines agency as ‘the capacity for self-willed action’. Entities capable of self-willed action should be described as subjects, in the sense that they experience their own world subjectively. We agree with the materialist agency approach that an important aspect of agency is that by the action the subject exerts an influence on the world around her. However, for an action to be more than mere behaviour, it seems that a desire or will should be behind the action. Note that an action that exerts influence and that expresses a desire or will does not presuppose the presence of intentionality, at least not intentionality in the strict second-order thought sense of the word. Consider the following example: a (human or non-human) animal can see a piece of food, say an apple, desire to eat that apple and then move to grab it. This involves the intention to eat the apple, but does not necessarily entail that the animal reflects on his or her own desire for the apple before acting to grab it. An agent can express her will or desires and influence the world around her without necessarily having the capacity to think about how the action will impact on others. Our working definition, which is inspired by Blattner, Donaldson and Wilcox (2020) and Sebo’s idea of perceptual agency starts from the idea that agency is the capability of a subject to influence the world in a way that expresses her desires and will. This capability springs from the phenomenology of the individual in question—her genetic make up, capacities, physicality and so on. In addition, and here we add a different layer to Sebo’s view of perceptual agency, it is important to realise that desires and wills do not come about in a vacuum. In the constitution of desires and wills, and thus also in agency, relationships—on the individual level, but also social, political and cultural structures—form an important role. In other words, there is always an interaction between agency and an animal’s environment. Social, political and cultural structures can limit animals’ capability of agency. While they might have the capacity for having agency, animals might be limited in their expression of this agency. Think of a zoo animal that is limited in her ability to roam freely. The animal, say a deer, may still have the capacity for running long distances, but in the zoo environment cannot exercise that capacity. On the other hand, zoo animals may develop new forms of agency in this limited environment, that they did not have before. The capacity for agency, can therefore be developed or hindered in interaction with the animal’s environment.Footnote 1 In this sense, Anthropocene conditions will also influence animal agency . In our view, in order to do justice to animal agency in our moral deliberations , we need a relational model that takes animals’ perspectives into account, as well as the social-historical context, and that does not measure other animals to a human standard.

3 Taking into Account Relational Agency in Animal Ethics on the Micro- and Macro Level

Acknowledging that animal agency matters and that we need to make space for their perspectives should lead to a relational, situated approach to ethics, in which not the human subject is the standard, but that focuses on the other. Drawing on insights developed by ecofeminists such as Carol Adams and Lori Gruen, we aim to move the question of how we should take agency into account ethically, past assessing the content of animals’ minds and building a judgment on that, to assessing the social context in which agency is curtailed or fostered. While we need more empirical research into animals’ minds (cognition and emotion) and cultures, we also need to focus on the social conditions that foster or constrain agency. This latter aspect is often underestimated in philosophy and animal ethics. This is problematic because the conditions of the Anthropocene target not only individual animals but also social groups and even species, so we need to acknowledge human responsibility not just in individual relations, but also on the macrolevel.

3.1 Relational Agency and Animal Ethics

(Eco)feminist approaches to ethics take (unequal) relations as the starting point for ethical considerations. Acting morally does not simply involve following rules, as in deontological approaches such as the animal rights theories mentioned above, maximizing happiness, as utilitarian approaches demand, or perfecting one’s character as virtue ethics requires, because the focus of our acts should not be on the self but on the other (Held 1990). This other is always a real, situated other, not a universal or idealized human. Ethical judgments are for this reason not universal but always tied to a context. Feminist ethicists emphasize that all of us are born into webs of relations and are dependent on others at several points in our lives, to different degrees. This dependency is not something to shy away from: relations with others are an ontological given, they constitute who we are, and can be a source of strength. As ecofeminists (Adams 2010; Donovan 2006; Gruen 2015) argue, the individuals we stand in relation to are not just human: we are also always entangled in relations with animals of other species, and they with us.

This relational and situated approach to animal ethics adds a different dimension to how we understand agency, compared to the ones sketched above. Agents are always tied to specific circumstances, capacities, and contexts, which influence their options for acting. Humans for example are born as a certain gender, in a specific culture, as a body that has certain cultural advantages or not, with a specific skin colour, in a specific class. They can choose professions, religions, partners, and so on; life will bestow hardships and joy onto them. The same applies to other animals. Social relations, physical dispositions, work and luck can all play a role in one’s options for exercising agency. Species characteristics matter, but are never the whole picture. We are furthermore all entangled in different relations with individual others that can create different forms of interdependence, influencing our agency and autonomy.

This way of conceptualizing agency in relation to ethics recognizes that there are different degrees of agency and intentionality, and that there is no one strict line between species when it comes to exercising these. It also shows that in formulating ethics it is not enough to just focus on biological capacities of certain species. Focusing on the social dimensions of relations between humans and other animals can for example help us see the role unequal power relations play and have played in their options for exercising agency. This is perhaps most clear in the case of domesticated animals, but the lives of humans and non-domesticated animals are often also intertwined. Think for example about the animals that reside in our gardens; our actions and theirs exert influence on each other. Understanding that human and nonhuman animal lives are entangled has a normative dimension. Lori Gruen (2015) argues that in order to do justice to others, including other animals, we should develop a caring perception that she calls entangled empathy. This form of empathy is focused on the wellbeing of others, and developing this entangled empathy is a process in which emotion and reason play a part.

Gruen’s entangled empathy is helpful in thinking about ethical relations with the individuals we encounter. However, most animals are also entangled in relations with others on social, cultural or political levels, and these entanglements often also strongly influence our scope for decision-making. This matters for ethical judgments. Individual nonhuman animals all belong to certain social groups, similar to humans, and this influences their scope for action. Legislation based on categorization of species as wild or domesticated determines space of movement or protection for individuals. Cultural constructions determine whether some city animals are seen as pests, such as rats (see chapter by Nieuwland and Meijboom in this volume), and others as belonging, such as songbirds, which has a strong impact on their options for self-realization. Cats who are born as companions have a very different set of options for agency than feral cats (see chapter by Meijer in this volume). Above we discussed the different specific harms that concern the agency of farmed animals. In order to adequately conceptualize agency, and to formulate new ethical guidelines, we also need to take these social and cultural aspects into account; in other words, we need to take the macro-level into account as well.

3.2 Taking into Account Macro-Relations in Thinking About Agency and Ethics

Recognizing the importance of social and political relations with other animals, Donaldson and Kymlicka (2011) developed a theory of political animal rights. They propose to view different groups of nonhuman animals as social groups based on their relation to human political communities. Specifically, they argue that wild nonhuman animal groups should be seen as sovereign nations, liminal nonhuman animals—those who live amongst humans in cities or rural areas but who do not desire close relations with them, such as mice, crows or feral rabbits—as denizens and domesticated animals as citizens. In all of these groups they emphasize the animals’ agency, arguing that they and not humans usually know what is best for them. For wild nonhuman animals, the good life usually means a life without human interference; for liminal animals this can involve contact with humans under certain conditions. For domesticated animals it often involves more contact with humans, because many of them need or desire some human assistance to flourish. This does not mean that they have no interest in shaping key aspects of their lives themselves, nor does it imply that the current power relation, in which humans are hierarchically above them, is just or unavoidable. Donaldson and Kymlicka argue that for these reasons domesticated animals, co-citizens in shared communities with humans, should have the right to be represented in political decision-making and should have the right to be included in the people in whose name the state governs: they should also have democratic agency, meaning the right to co-shape common decisions.

With their theory of animal rights they draw attention to nonhuman political agency, which is often overlooked and erased (Meijer 2019), using examples that range from resistance to co-creating common interspecies communities. They also draw explicit attention to the distinction between micro-agency and macro-agency, arguing that macro-agency is also relevant in the nonhuman context (see also Donaldson and Kymlicka 2013). Micro-agency refers to the scope for making personal decisions, which can for domesticated animals include deciding on what to eat, where to sleep, and who to play with in the park or garden. Only focusing on this type of agency, however, obscures how larger scale power relations, reified in laws, institutions, and political and social processes, form the scope for this micro-agency. Dogs can in many places in the world usually for example not decide to leave the house they live in when they desire, even if their human would allow them to, because cities are not safe for them and there are leash laws and areas where they are not wanted. Horses cannot choose to leave their meadow or barn and are usually not consulted before riding, even when some riders take their preferences into account in deciding whether to engage in dressage, jumping, or other sports (see Meijer 2019 for a longer discussion of this problem). Wild nonhuman animals have no say in the preservation of their habitat. Donaldson and Kymlicka show that taking animal agency seriously implies more than simply reformulating our individual relations with them: humans should also consider their political institutions and processes, laws, and rights, to incorporate nonhuman animal perspectives.

Especially in the Anthropocene taking macro-agency into account is of the utmost importance, because the context shapes how and sometimes even if animals can exercise agency. As we mentioned above, in the Anthropocene animals’ agency risks being increasingly curtailed due to upscaling of livestock production, habitat loss and fragmentation, and climate change. In our globalized world, the conditions for our relations with other animals are institutionalized in many ways. Economic, political, legal and cultural structures determine their and our space for movement and for creating new relations. Developing an animal ethics for the Anthropocene thus asks for more than considering our individual duties: it also implies carefully rethinking political and social institutions and practices. In the context of interacting with other animals this means taking into account these macro-factors, in epistemic and ethical judgments. But it also has ethical implications beyond that, such as that we should aim to change these large-scale oppressions.

4 Risks for Relational Approaches to Ethics

While a relational approach opens up new ways of engaging with other animals and co-creating communities, there are also risks. One familiar argument against relational approaches is that they are too dependent on context, thereby making it impossible to formulate clear ethical guidelines. Following this, ethical judgment is not universal, but relative to communities. In the case of animal ethics this could lead to unequal treatment, but also even mistreatment of animals, in particular in contexts where the human-animal relationship is an instrumental one, such as in industrialized farming or animal experimentation. We recognize this risk. We understand agency to be important, because it enables subjects to shape their own lives, in line with their subjectivity. This presupposes a certain view of the animal subject, and is part of respectful engagement with them, which is not compatible with killing or abusing them for human benefit. We believe that this has been adequately argued for by animal philosophers before and therefore focus on the next step of the argument (see also Donovan’s 2006 reply to such criticism). Arguing that context forms precise secondary rights and duties also does not have to be random: we are in favour of developing clear outlines for this project. As it involves dealing differently with other animals, however, these cannot be provided beforehand; after all, the outcomes are dependent on the input from other animals.

A second risk involves anthropomorphism, falsely attributing human characteristics to other animals. While all theories of animal ethics begin with a human framework to assess animals’ standing, relational theories seem to require more interpretation, because they involve communication, a focus on a changing context, and a place for animal perspectives in social and political decisions. Furthermore, even seeing animals as agents in these relations is according to some (compare Sebo 2017) a matter of anthropomorphism. To begin with the second point: not attributing any emotions or cognitive content to other animals is not a neutral stance, but rather the outcome of (Western) power relations. De Waal (1999) calls this ‘anthropodenial’. It is a self-serving ideology, because it closes off animal participation beforehand. While we are not sure how relations with other animals can evolve, we do note that if we presuppose they have no agency and are not capable of new forms of interaction, these new forms will never happen. This relates back to the first point. Relational approaches do not require more interpretation. They require a different form of interpretation. Instead of taking a view from nowhere, and from there once and for all defining what animal expressions mean, our aim is to learn more about animal agency from the ground up, in human/non-human animal interactions, step by step. In this process philosophy, ethology, and other fields of study have a role to play, as do actual relations with nonhuman animals. While risks of interpretation and context are something to be aware of in relational approaches, they are inherent in doing animal ethics more generally. As such, they should be given attention, but they should not be overstated.

A final issue for consideration is the relevance of species membership. Focusing on relations might give the impression that biology and ethology no longer have a role to play in animal ethics. It would simply be a matter of social or political philosophy, and experimenting with new forms of co-habitation. However, ‘animals’ are a heterogeneous group. While we know quite a lot about certain animal species—such as dogs or dolphins—about most species we do not know so much. We therefore need more empirical research, as we will explicate in the next section, on species who are very different from humans, and on those who live close to humans but who are usually not seen as important research subjects for their own sake, such as farmed animals. In order to be able to take specific animals’ perspectives more seriously, we need more, and different kinds of, knowledge about the species the animals belong to.

5 Further Directions

A relational ethics for the Anthropocene should take into account nonhuman animal voices and perspectives. According to feminist standpoint ethics, the voices of socially suppressed groups need to be heard, as ‘their views are found inevitably to be subversive of the ideological system that would render them silent – sexism in the case of women and girls and speciesism in the case of animals’ (Donovan 2017, 210). This implies taking responsibility for human actions while at the same time acknowledging the influence and acts of nonhuman animals. This asks for an attitude of empathy, listening and curiosity, both with regard to attending to actual nonhuman animals, and with regard to locating injustice and domination in existing larger scale relations. This might sound utopian, but we already find examples of new relations between humans and other animals. As a conclusion we will highlight these in three different fields: animal research, cultures, and work.

5.1 Research

Models of animal research are traditionally human-centred. There is a hierarchy between the human researcher and the animals, both materially—animals are often kept in cages—and epistemologically—animals are not seen as interlocutors but as objects of study. Ethologists such as Smuts (2001) and Bekoff (2007) challenge this model, studying nonhuman animals as subjects, and not as objects. This leads to new methodologies—such as following them in their habitats, and letting them co-shape the conditions of the studies. Smuts for example describes how she had to learn to ‘speak baboon’ in order to be able to study a group of baboons. Scientists usually try to ignore primates, so as not to let their presence near them influence their interaction (Smuts 2001). Smuts found out that ignoring the baboons was not a neutral act, because baboons are social animals: she had to learn to interact with them on their terms. Furthermore, by interacting with them, she experienced critical aspects of their society, such as hierarchy, personal space and communication, directly. Because the questions asked determine the answers nonhuman animals can give, adopting this kind of attitude towards animals matters greatly to learning about their inner lives and cultures, leading to knowledge that can and should inform ethical theory. Taking animal agency seriously most likely also implies not using them in forms of experimentation that infringe on their liberty or harm their welfare.

5.2 Animal Cultures

Another example concerns the increasing recognition of the importance of animal cultures. In recent years animal culture has become a topic of study in many species. Elephants are perhaps the most famous example. They usually travel along the same routes every year (Barua 2014). Knowledge about these routes and specific locations, for example where water can be found in dry seasons is transmitted culturally. Matriarchs teach the younger elephants the ropes of survival. Human activity increasingly disrupts these patterns of knowledge exchange and other cultural processes, in several ways. Humans may designate a certain area as a nature reserve, and use fences to close it off—sometimes with the best of intentions, aiming to protect the nonhuman animals living there—which makes it impossible for elephants to find their way. Poachers also often kill matriarchs and other older elephants in a group, which traumatizes younger elephants, and makes it very difficult for them to rebuild their lives afterwards. Gay Bradshaw researches PTSD in elephants and shows that while intensive human care may help younger elephants to recover from the trauma’s they witnessed, it is very difficult for humans to take on the cultural role (Bradshaw 2009). When the matriarchs die, often that specific cultural knowledge dies with them.

Other examples of cultural knowledge being transmitted concern the migration of bighorn sheep, birdsong, and chimpanzee fashion. Conservationists take this cultural dimension increasingly seriously (Laland and Janik 2006), but respect for and knowledge about nonhuman animal cultures is not only relevant for conservationists. It is important for politicians who design new legislation for liminal animals, city planners, animal rescue organizations, and others dealing with groups of nonhuman animals, their travelling routes and habitats.

5.3 Animal Workers

New relations are also found in the context of work. While most nonhuman animal workers, for example in factory farms, are exploited, there are also forms of work that can benefit both human and nonhuman animals. An example concerns crow workers. Dutch start-up, Crowded Cities, plans to train city crows to pick up cigarette butts. Using crows as cleaners raises many questions about their working conditions, but when the work is safe and their freedom is not compromised, this working arrangement could be beneficial for crows, humans, and the environment. At this stage, more research should be done on the benefits and burdens of this type of work for them, including monitored pilots of the project, in order to make sure they are not exploited.

Another example of animal work concerns domesticated support animals, such as for example rescued dogs who go to hospitals to distract young patients. The benefits of therapy animals for humans have been proven scientifically (Glenk 2017). The health of these animals, and the possible benefits for them, have however not been studied in detail (ibid.). Many companion animals suffer from boredom and many like to work, so for them working could contribute to better health and happiness. In order to establish which types of engagement are possibly beneficial for humans and other animals, we need more research into the benefits and burdens of care work for nonhuman animals. This could for example lead to establishing labor rights, including the right to play, time off to do stuff with friends, rest, and a pension when they are old (Cochrane 2016).

5.4 Further Directions

These three examples show that establishing an ethics for the Anthropocene is not a matter of all or nothing: ethical relations exist and existing relations can be improved. Furthermore, they show that developing an ethics for the Anthropocene should be an interspecies project. In order to give shape to an animal ethics for the Anthropocene we first need to further develop our understanding of animal agency ‘from the ground up’, through studying animal behaviour and interspecies relationships.Footnote 2 We also need to further reflect on the question of how phenomena like agency, intentionality, autonomy, and self-realisation relate to each other, and how they will change once we take seriously the different ways in which other animals relate to these concepts. Animal ethics is not something to be thought out solely by humans behind computers, it is something for which humans also need to engage differently with other animals. This is important for several reasons. We need more, and different forms of, empirical research, to find out the scope of their agency, their view on relations, and to find out how we can build better relations, that are beneficial to all those involved. It is also important to decentre the human and find out how we can theorize together with other animals. Perhaps it even requires a move from animal ethics to an interspecies ethics, at least with regard to the (domesticated and liminal ) animals with whom we share our lives, households and cities.